All articles · 6 min read

Why do two identical windows measure differently on my floor plan?

Two windows that look identical in real life can easily end up with different measurements on a surveyed floor plan. It is rarely the windows that differ — it is the measuring itself.

Where does the difference come from when the windows look identical?

An opening in a wall — window or door — is measured by something (a sensor, or your own eye and tape measure) finding the edge of the hole in the wall. That edge is not a mathematically sharp line. There is plaster, a slight rounding, shadows, and for a scanner: a spread in exactly which point counts as "the edge". Do that twice on two different windows — even identical windows from the same batch — and you get two independent readings, and two independent readings rarely land on precisely the same figure.

The difference is usually small: a few millimetres up to a centimetre in width or height. That is below what you can see with the naked eye, but it is enough that a window schedule with two rows that should match shows "118 cm" in one row and "120 cm" in the other.

Could it also be that the windows really are different?

Yes, that happens — and it is worth checking before you assume measurement noise. Two windows that look alike from a distance can have been fitted by different tradespeople at different times, or one may have been replaced at some point after construction without being an exact copy of the original. A difference of a few centimetres can therefore be real.

Rule of thumb: if the difference is under 1-2 cm, it is statistically measurement noise. If the difference is 5 cm or more, you should consider whether the windows are actually different types or sizes, and check with a tape measure before correcting anything.

What is sill height, and why is it mentioned here?

The sill height is the height from the floor to the underside of the window frame — in other words, how far up the wall the window starts. It is a third measurement alongside width and height, and it is just as exposed to the same kind of measurement noise: two windows in the same room, sitting at the same height, can easily end up with a sill height that differs by 2-3 cm on the drawing, even though in reality they sit at exactly the same level. If you need the full explanation of what this measurement means and what is typical, there is a separate article on exactly that.

Why is this even a problem worth fixing?

Most of the time it is cosmetic — a table where two rows don\'t quite match looks untidy but changes nothing in practice. The problem becomes real when the figures are used for something: a curtain maker cutting two curtains to two different widths when they should really be identical; a window supplier receiving two different order measurements for a pair of windows that should match; or a window schedule you pass on, where the recipient has no way of knowing whether the difference is a real detail or just noise.

How do you fix it without guessing?

The robust solution is not to average the two measurements or pick the "neater" figure — that is still a guess. The robust solution is to decide which of the two measurements you trust most, and use that same figure in both places. If you have checked one of the windows yourself with a tape measure, the tape measure\'s figure wins. If both measurements come purely from a scan and you are confident the windows are identical, you can copy one measurement onto the other rather than leaving two independent, noisy figures standing side by side.

Copying is faster and more reliable than manually rounding each figure to a neat number — a round number is still a guess, whereas copying from an opening you trust is at least internally consistent.

The width is copied symmetrically — why does that matter?

When an opening\'s width is changed, there is a choice to make: which side of the opening moves? If the width changes symmetrically around the opening\'s centre, the centre is preserved and both edges move equally. That is usually the right choice for a window, because a window is typically centred relative to something — a piece of furniture below it, a fixed position in the masonry, or simply the eye\'s expectation that a window sits centred within its bay. If instead only one edge moved, the window would shift within the wall, which is rarely what you mean when you\'re just correcting a figure.

How do you avoid the problem from the start?

• Scan slowly and close to the openings. The closer the sensor gets to a window frame, the less spread there typically is in the edge detection.
• Check the important measurements with a tape. Windows and doors that will be used for ordering — curtains, frames, new panes — should always be checked manually, no matter how good the scan is.
• Note which windows should be identical while you\'re standing in the room. It\'s easier to remember on the spot than to guess afterwards from a drawing.

When should you NOT assume the windows are identical?

In older houses, where extensions and replacements have happened over decades, it is common for windows that look alike from a distance to genuinely have different measurements — a centimetre or two in width, a different sill height. If you need to order something that must fit precisely, it is never excessive to bring the tape measure and check, even when the drawing shows two identical figures.

What HouseSense does

If two openings have been measured separately and should share the same measurement, you can copy the width, height and sill height from one to the other. When the width changes, it is kept symmetrical around the opening\'s centre, so the window doesn\'t shift within the wall. This doesn\'t replace checking with a tape measure if the figures are going into a binding order — it simply corrects two independent, noisy figures to agree with each other.

In short

1. A difference under 1-2 cm between two "identical" windows is usually measurement noise.
2. A difference of several centimetres can be real — check it before correcting anything.
3. Never blindly round to a neat figure — use the measurement you trust most.
4. If the measurement is going into an order, a physical tape measure is still the final check.